+- Gekko Forum (https://forum.gekko.wizb.it)
+-- Forum: Gekko (https://forum.gekko.wizb.it/forum-13.html)
+--- Forum: Strategy Development (https://forum.gekko.wizb.it/forum-12.html)
+--- Thread: Settings between backtest / live (/thread-57554.html)
Settings between backtest / live - Guysmo - 08-13-2018
Hi,
We use the Backtest tool to know if our strategy is profitable on a long run. I read that Tommiehansen and Jvs are doing millions of tests to find what are the best settings but recently I saw we can modify parameter's value on the fly to be as close as possible to the market trend, in order to increase the profit. But if we change parameter's value, that's mean the backtest result is wrong compared to our current strategy ?
Thank
RE: Settings between backtest / live - askmike - 08-13-2018
Quote:But if we change parameter's value, that's mean the backtest result is wrong compared to our current strategy ?
Old backtests yes, but you can run a new one (with your new parameters) in a second. I've just added some docs to the backtesting documentation that might be of interest. See here: https://gekko.wizb.it/docs/features/backtesting.html
RE: Settings between backtest / live - Guysmo - 08-13-2018
(08-13-2018, 09:10 AM)askmike Wrote:Quote:But if we change parameter's value, that's mean the backtest result is wrong compared to our current strategy ?
Old backtests yes, but you can run a new one (with your new parameters) in a second. I've just added some docs to the backtesting documentation that might be of interest. See here: https://gekko.wizb.it/docs/features/backtesting.html
Thanks for the article.
If I understand we can't keep in mind our backtest result will be always good. So I guess the best way is to try on paper trade while adjusting parameters on the fly with bull / bear / sideways market ?